3D Model Pricing

by Matt Wisdom

I wanted to take a moment to discuss some of the issues around 3D model prices, and whether the price per model is worth the effort for some artists to publish and sell. Many artists make their entire income from TurboSquid, and we’ve spent an enormous amount of time studying this topic, so I wanted to share some thoughts. I’ll start with the conclusion — the answer for artists who want to sell at higher price points is to give customers more value for their money at those higher prices, with less risk and hassle. However, the road TurboSquid took to get to this pretty obvious conclusion wasn’t so obvious.

When we started TurboSquid, pricing was low. We were competing against Viewpoint, who charged US$1000 per model. The prices artists set in a free market like TurboSquid were low enough that we worried there wouldn’t be a high enough volume (number of individual sales) to justify a low price. Viewpoint had done an economic study, and for those of you who follow this kind of stuff, they determined that people only bought models when they were desperate, and were willing to pay high prices, what is called “inelastic” in economics terms. So Viewpoint created pricing that was so high, people truly only bought when they absolutely had no other choice. In this way, the results from their economic study were self-fulfilling.

We created TurboSquid out of broader ideas for the Internet in general and digital marketplaces in particular, and since our founders were 3D artists, 3D was a great place to focus. We had lower pricing and a wider variety of models, and for better or for worse, a wider variety of quality than Viewpoint did. If we were airlines, customers buying Viewpoint weren’t just traveling first class, they were paying for the Concorde at stratospheric prices. TurboSquid was the Southwest or Ryan Airlines of that comparison. Viewpoint is almost forgotten, except for us old timers who remember what we used to like about their catalog back in the day.

After we’d had a marketplace going for two years, TurboSquid decided to raise prices in 2002 because prices were so low we were concerned we were short-changing both ourselves and our artists. We did a mass price increase with strategically-chosen 3d models to about triple their previous prices. The result was that we lost volume, but revenue stayed about the same. In other words, we traded a lot of customers spending less for fewer customers paying more, which is generally considered to be a bad trade. However, it did disprove Viewpoint’s conclusions–the trade proved that prices are indeed “elastic.”

We did a lot of analysis right after the change, and then dropped some prices to try to restore some equilibrium. In any case, it was very difficult to tell which prices should be higher or lower — how to find an “optimal” price that maximized revenue seemed impossible with the data, and we had outsiders and PhD’s review the data as well. As a general process going forward, we then had somebody review each new model to enforce the higher pricesĀ and look at prices for models under related searches, and then change the product’s price if it was too low, or more rarely, too high.

After a time, it really felt like we were just injecting noise into the marketplace. Why had the initial raised price been selected? Was it right? Did prices go too high initially and we were merely keeping that mistake going? Too low? The data were just too sparse to come to solid conclusions. So around 2005 we unlocked pricing completely and let the market set prices.

Interestingly, the average selling price of a model correlates to publishing rates. Throughout the history of the company we’ve seen the publishing rate per month increase. More and more models are published each month, which seems to correlate to a slow drop in the prices of models sold, continuing over time. It was hard to detect at first, but over years, very clear. It seems like this was caused more by increased supply at various price points, than a general lowering of pricing from artists. Given various price points, customers were trending lower in the amounts they would spend on a given model. But they kept on buying more and more volume, and the overall sales revenue of the marketplace continued to improve. It turns out TurboSquid would’ve succeeded anyway without the price hike.

The financial crisis of late 2008 caused both a spike in publishing as artists looked to beef up their income from stock 3D, and a drop in average price per model sold as customers tightened their belts. Clearly this drop in average pricing was caused by customers choosing to spend less money — our analysis tells us that artists didn’t just drop their pricing en masse all of a sudden. Customers started to buy models again in higher volume in late 2009 and have been buying ever more since. The average selling price over the last year has been about double the original pricing of 2000-2002, having drifted down from triple that original pricing.

Looking to the future, there are a lot of benefits if the average selling prices rise for models. TurboSquid would like to see this happen, and remember, we went to the mat before fighting for this in a storm of criticism. But we don’t just want to see prices rise for the sake of taking more money from customers per model. We did that before because we felt we had to, and all we did was trade to fewer customers. The trick is really this — the models have to have significantly more value for the prices to rise. Like the airline analogy, the customers that want to fly business class want models that work with no headaches. They want to save time, make their deadlines with less stress, and are willing to pay more for that value. We’re not going to override the market in setting prices above what artists want to set ~ we’re going to add value to products, and customers will see that and, we believe, choose the premium products, but it’ll be their choice to spend more money.

So everything circles back to TurboSquid’s ability to invest in the models and figure out what customers need, and then work with the artists who do the bulk of the work to create something of clear value that customers *want* to pay more for. After two months of CheckMate, we have the first significant uptick in average selling prices for models across the whole marketplace since we let the market set the prices. It’s very clear when you see the graph. We said internally at the outset of the program that if CheckMate is a success, the trend of the price points for selling models will go up for our marketplace as a whole. That’s really the only statistically significant data because once a model is CheckMate certified, it’s hard to know how it would have sold if it hadn’t been certified, and we’re seeing the hoped for increase in these selling prices.

I think if you asked the other staff members, they’d tell you that CheckMate has been a success beyond our wildest hopes. In a broad sense the numbers are really astonishing, with relatively few models having an outsized impact, and drawing customers that had a bad model quality experience before to come back. We intend to protect the CheckMate price points — these models are exclusive, they aren’t labeled CheckMate anywhere else, and we will keep the prices at a level where artists want to invest in the models and customers know they can get that value reliably at TurboSquid. Models should sell, and particularly CheckMate Pro models, at a higher price point. Now that we know TurboSquid’s capacity for inspecting models, we are going to introduce new price points so more CheckMate Pro models can be certified at lower prices when they are simpler models. Complex Pro models will be higher priced, period, but we will have premium Pro models at a variety of price points, some lower than now.

Going forward, we have some very good ideas about how to be much more specific about pricing, and add the strength to the analysis of pricing that we were frankly missing the last time we tried to improve model pricing. This is going to take some months to build out to the rigor we need, and we are putting a ton of energy into it. We’ll probably even ask for your help. It’s not a trivial problem that can be solved by having an individual simply sift through models, as we’ve seen already. The problem is much more complex when you consider the many varying categories and levels of complexity in geometry, texturing, accuracy, aesthetics, etc.

A question: should TurboSquid let artists on the marketplace that sell at lower prices? The answer is yes, because we don’t feel like we can solve that problem for the general catalog of models — there will be a marketplace whether we host it, or it is scattered on the net. Without question, a great model for $1 is hurtful to the ability of artists to invest in creating similar types of objects. It’s a category killer, and customers could think they’re stolen, and we don’t want that. But some artists have collections at relatively low prices that are genuinely trying to maximize profit — they’re not throwing models out there to damage the marketplace, which is a huge distinction.

Maybe that’s why it feels so threatening to see the value at those prices from some artists, because it is a high quality organized effort to produce quality models that are very competitive. The thing is, I can name some artists that are targeting architects who have lower budgets than most 3D projects and need collections at low price points. The solution from our point of view is not for TurboSquid to get back into raising prices on models, which would have the bad side effect of making TurboSquid arbitrarily more expensive than other outlets on the web, or to start kicking models off the site from professional businesses who trying to make a real living selling models.

The real value we see — and the money for artists — is in giving customers better value for *their* money, through higher quality models with less risk. This is where CheckMate is headed, and where so many customers who see tons of value in stock 3D want all of us to put our efforts. Looking at the true low end of the market, there is a never-ending stream of free junk coming from Google’s 3D Warehouse. We can’t, and shouldn’t, compete with that.

Our strategies go beyond CheckMate. We have multiple approaches to try to get customers the most benefit possible, where customers are willing to spend way more money on stock 3D, and ultimately live more creative lives because they don’t “reinvent the wheel” by building models readily available at the expense of their creativity elsewhere. Artists who publish similarly want to live more creative lives, with less risk and more financial stability. TurboSquid’s duty is to try to maximize value for everyone, and we spend way more time trying to solve these problems than I ever believed possible back when I was an animator. And as best we can tell, the long term solution is to build around CheckMate as the foundation, with the support and deep analysis to keep the value and pricing at that magical intersection of demand, volume, and price which artists and customers consider “optimal”.

I’ve traded my life’s work from something directly creative like 3D animation and compositing, into something heavily analytical, financial, and managerial. It’s been a good trade because I believe in what we’ve all created in TurboSquid. I try to solve these problems around pricing and quality by leading our team and devoting an enormous amount of energy and thought to them. CheckMate is the best answer we’ve found, and I deeply hope the trend toward increased selling prices continues because we’ll all invest more, and our customers will be increasingly delighted with the value and quality of what they purchase here. I hope this helps everyone understand the history, and our feelings about pricing, marketplaces, and the evolution of quality in the stock 3D world.

Tags:

3 Responses to “3D Model Pricing”

  1. Great post Matt,
    I’m lucky enough to have been with TurboSquid through the pricing changes and I have to say that I admire the attention you guys give the topic. It’s a tricky problem to solve, and I believe that Checkmate is a great tool to sort of, manage the issue. I’m excited to hear about lower priced Checkmate Pro models as I’ve got a few models that could be Pro rather than Lite if the price minimum didn’t seem so disproportionate for the model in question.

    Recently, I was working on a project that required a model of a snowmobile. I knew that modeling one was way outside of the time-scope of the project and I searched on both TurboSquid and Google. I found the same model on TurboSquid and another site (This was Pre-Checkmate) and although it was cheaper on the other site, I decided to buy from TurboSquid. When I made that decision (based mostly on the fact that TS support is very good) it was kind of an eye opener for me. I was now the customer and not the artist… It was quite strange indeed.

    I’ve recently been moving to certify (almost) all of my models not for the badges, but for the idea that a customer might relate to ths the way I related to the customer support.

    You guys are doing some good stuff for the industry. Keep it up!

  2. Matt Wisdom says:

    Thanks, Jamie. It’s been a fascinating process to learn about all of these topics, and even how things like pricing change in the internet age with downloads and infinite supply. Someone is going to create a new branch of economics about how all of this, virtual worlds, virtual currency and related stuff change the assumptions of basic economic theory. I wish I could read the future textbook on the subject now!

    We’re seeing over and over again that customers care a lot about getting good solid quality with no worries. The fact of TS being a marketplace is much less important than the products themselves, no matter where they come from. We used to emphasize how big our selection is to imply how many types of things a customer could find, but that’s not important anymore because the library is so big. For us, it’s all about quality now.

  3. Basas says:

    I am just starting with TS, or should I say, considering to make 3D modelling as my main source of income, and am really glad Matt took his time to explain it all. Because after going through the forum complaints about undercutting I was already very sceptic on how reliable this choice of earning is.
    I ended up on a forum trying to find some statistics on how getting your product CM certified increases the sales, or some statistics on percentage for viewed/bought products. It would help allot to be able to compare these things to the weighted average of the whole site, only same category, same price range etc., etc. In short, what I am suggesting, that sharing more of statistics with your vendors, and giving more tools to do statistical analysis, would help auto regulate the price/quality more accurately and so much faster. In other words, all the tedious work you do analysing your data and looking at the big picture could be done by sellers individually and adding to your cause. Maybe this will help the new sellers more rather than the old-timers who already have what to compare things to, but I doubt many of them export their statistics every so often to compare it and to look at the graph.

    I’d like to thank you for your great work, I’m only starting to take this place seriously, but I already feel like at home. It feels like TS with it’s all staff are here to make a difference in a field they all care about, and the “getting paid for what you do” part only comes as a by-product ;)

    Regards

    Basas a.k.a. Storm

Terms of Use Privacy Policy Site Map © 2013 TurboSquid